There has been much interest of late in the production of complexes that involve partially unfolded proteins and lipids. These proteins may have drastically different properties and particularly biological properties than the corresponding proteins in a fully folded state. The gain of new, beneficial function upon partial protein unfolding and fatty acid binding is a remarkable phenomenon, and may reflect a significant generic route of functional diversification of proteins via varying their conformational states and associated ligands. Thus, in addition to alternative splicing of mRNA transcripts, post-translational modifications and changes in tertiary structure of specific domains, partial unfolding of a previously native protein is becoming recognized as a mechanism to generate functional diversity. This may be due to a cellular response to unfolded proteins and to the lipid cofactor, which defines their altered properties. However, this response may be different in for instance tumour cells, which means that they may give rise to therapeutic potential. In order to form stable moieties, the unfolded proteins are frequently modified in some way, and in particular may be bound to cofactors such as fatty acid cofactors. The complexes formed in this way may be stable and give rise to therapeutic options.
HAMLET (human alpha-lactalbumin made lethal to tumor cells) is one such example of a new family of tumoricidal molecules, with remarkable properties. Formed from partially unfolded α-lactalbumin and with oleic acid as an integral constituent (Svensson et al., 2000, PNAS 97: 4221-4226), HAMLET was discovered by serendipity when studying the ability of human milk to prevent bacteria from binding to cells (Hakansson et al., 1995, Proc. Natl. Acad. Sci. USA 92: 8064-8068). Early in vitro experiments showed that HAMLET displays broad anti-tumor activity with a high degree of tumor selectivity and subsequent therapeutic studies have confirmed HAMLET's tumoricidal activity and relative selectivity for tumor tissue in vivo. In a placebo controlled clinical study, topical HAMLET administration removed or reduced the size of skin papillomas (Gustafsson et al., 2004, New England Journal of Medicine 350: 2663-2672) and in patients with bladder cancer, local instillations of HAMLET caused rapid death of tumor cells but not of healthy tissue surrounding the tumor (Mossberg et al., 2007, Int. J. Cancer: 121: 1352-1359). Therapeutic efficacy of HAMLET in bladder cancer was recently demonstrated in a murine bladder cancer model (Mossberg et al., 2010, The Journal of Urology 183: 1590-1597), and HAMLET treatment delayed tumor progression and led to increased survival in a rat glioblastoma xenograft model without evidence of cell death in healthy brain tissue (Fischer et al., 2004, Cancer Research 64: 2105-2112). HAMLET thus appears to identify death pathways that are conserved in tumor cells, thereby distinguishing them from healthy, differentiated cells.
Complexes using equine lysozyme and oleic acid have also been found to produce cell death (Vukojevic et al., Langmuir, 2010, 26(18) 14782-14787), suggesting that different, unfolded proteins can become cytotoxic when coupled to a suitable cofactor.
Ion channels are membrane proteins that sense alterations in membrane tension or cellular environment, and their activation has been proposed to catalyze several signaling cascades, including ER stress, p38 MAP kinases, small GTPases, PI3K/Akt and NFkB as well as Ca2+ dependent pathways, mainly in sensory cells or muscle cells involved in mechanical responses.
Channels are present in most cell types, including bacteria, however and recently, ion channel perturbations have been proposed to promote malignant transformation, tumorigenesis and metastasis (see for example Huang, S. et al., 2005, Cancer Cell 8, 175-176; Wolf, K. et al., 2007, Nat Cell Biol 9, 893-904, and Arcangeli, A et al. 2009, Curr Med Chem 16, 66-93) suggesting immediate relevance of such channels for cancer cell homeostasis.
Specifically, the applicants have identified one example of a molecule that acts as an ion channel activator with tumor specificity and propose that the broad death response of tumor cells and certain bacteria to this and other agonists with similar target specificity involves the perturbation of ion channels, which are conserved throughout evolution. The de-differentiation of tumor cells may thus involve the reversion to a more “primitive” ion channel repertoire, which may be targeted by such agonists. This tumor-selective death through ion channel perturbation is particularly relevant, especially in view of the molecules already documented protective effects against tumors in patients and animal models.
The applicants investigated the activation of ion sensitive and if signaling triggered by such channels might distinguish the death response of tumor cells from the survival response of healthy, differentiated cells. Rapid Na+ and K+ fluxes followed by mobilization from of intracellular Ca2+ stores was detected in carcinoma cells.
Inhibition of cell death by Amiloride and BaCl2, which block Na+ and K+ fluxes, suggested that death is triggered through the combined activation of mechanosensitive channels and potassium channels and inhibition of ER stress induction and the p38-dependent death response suggested that ion fluxes directly activate downstream signaling pathways that execute carcinoma cell death. Healthy, differentiated cells, in contrast, showed a weak and transient Ca2+ response under similar treatment and but no p38 activation and instead, an innate immune response accompanied their survival. It is possible that tumor selectivity in vivo may thus rely on ion channel perturbations and a p38 MAPK death response, accompanied by a beneficial innate immune response in surrounding tissues.
Defective ion channel signaling deregulates mechanisms of cell-cycle control, DNA-damage repair, apoptosis, adhesion and migration (Huang et al., 2005, Cancer Cell 8, 175-176; Wolf et al., 2007, Nat Cell Biol 9, 893-904). The relationship of ion channel function to cancer has therefore received increasing attention, and ion channels are becoming established as modulators of signals that promote oncogenic transformation. Understanding of ion channel aberrations in cancer progression is therefore essential and controlling their function may constitute an important new approach to cancer therapy. Targeting of ion channels in cancer cells has been proposed as a future therapeutic option (Arcangeli et al., 2009, Curr. Med. Chem. 16, 66-93) as has the control of mechanosensitive channels and other ion channels, which are overexpressed in carcinoma cells. Despite this proposed usefulness, the therapeutic potential of ion channel modulators remains underexploited, due, in part, to side effects reflecting lack of tumor specificity.
The applicants have now identified substances with ion channel activator activity with tumor specificity and propose that the broad death response of tumor cells and certain bacteria can involve the perturbation of ion channels, which are conserved throughout evolution. The de-differentiation of tumor cells may thus involve the reversion to a more “primitive” ion channel repertoire that is targeted by these substances.
While oncogenic transformation and cancer cell function require ion channel support, ion channel variability and complexity is considerable. Ion channel-encoding genes are frequently over-expressed in human cancers, due to gene amplifications, epigenetic regulation or splice variants of channel encoding genes but except for KCNRG, encoding a K+ channel-regulating protein with tumor suppressor properties, tumor-specific mutations in ion channel genes have not been reported. In addition, though most human cancer cells show altered Ca2+ wave dynamics, cancer-specific alterations in the “spatio-temporal nature” of Ca2+ waves or ion channel expression profiles have largely not been identified (Arcangeli et al., 2009, supra). In specific cell types, siRNA mediated inhibition of individual ion channels has been found efficient, but due to the complexity, knockdown of individual channels is often insufficient to obtain loss of function and to reproduce a phenotype relevant for cancer. The use of pharmacologic channel inhibitors therefore remains crucial for defining the general involvement of different functional classes of ion channels, even though each inhibition does not fully define a specific channel type.
Within these technical limitations, our results show that it is possible to perturb tumor cell membranes, leading to ion fluxes, depolarization and the opening of ion channels in such a manner as to discriminate tumor cells from healthy, differentiated cells. Resulting tumor cell death and morphological changes were shown to be ion channel-dependent, using pharmacological inhibitors and a link between ion channel activation and cell death was suggested by genome wide transcriptomic analysis, showing that channel blockade reduced the number of differentially expressed genes in treated carcinoma cells from about 400 to 40. Transcriptional regulation of the top scoring ER stress, p38, and Ras pathways was inhibited by the channel blockers, as was the phosphorylation of corresponding proteins and inhibition of ion channel activation prevented carcinoma and lymphoma cell death.
Our recent studies in artificial vesicles and tumor cell membrane models suggested that mechanosensitive channels are opened by treatment, as well. Mechanosensitive channels are gated by lipid bilayer deformation forces arising from local or global assymetries in transbilayer pressure or in bilayer curvature. Fluorescence imaging showed that an accumulation of the administered substance in receptor-free phospholipid membranes and perturbs their structure by elongation. Similar results were obtained with plasma membrane vesicles from tumor cells, which formed tube-like membrane invaginations after substance exposure. Furthermore, the applicants have found during some treatments, transient pores form in artificial lipid bilayers at physiological pH, possibly explaining the observed membrane leakage. Thus, in addition to ion channel activation, direct permeabilisation of carcinoma cell membranes might activate cell death.
The signaling profile in carcinoma cells was consistent with patterns previously observed after physiological activation of mechanical, stretch-induced channels in a variety of cell types. Mechanical membrane perturbations have been shown to perturb ERK1/2 via G proteins and especially p38 signaling (Correa-Meyer et al., 2002, Am. J. Physiol. Lung Cell Mol. Physiol. 282: L883-L891). JNK and p38 are also key mediators of signals stimulated by various stresses and are mainly responsible for responses such as stress-dependent apoptosis and inflammatory responses. HAMLET shifted the MAPK signaling profile of tumor cells from the ERK1/2 to the p38 pathway, and thus from proliferation to death. In mammals, MAPKs are divided into three major groups, ERKs, JNKs/stress-activated protein kinases, and p38, based on their degree of homology, biological activities, and phosphorylation motifs (Cobb, 1999, Progress in Biophysics & Molecular Biology 71, 479-500). MKK3 and MKK6 activate p38 MAP kinases by phosphorylation at Thr180 and Tyr182 and activated p38 MAP kinases phosphorylate and activate MAPKAP kinase 2 and phosphorylate the transcription factors ATF-2, Max and MEF2. Subsequent phosphorylation of p53 and CHOP, among other targets, leads to the activation of cell death mechanisms, including mitochondrial permeabilisation, caspase activation and DNA fragmentation, which has been shown to occur in some treated carcinoma cells. In addition, Hsp27 phosphorylation mediates cytoskeletal rearrangements, potentially explaining the change in morphology that we observed in carcinoma cells exposed to some treatment.
Changes in cytoplasmic Ca2+ concentrations may compromise the ability of the ER to correctly fold proteins, thereby eliciting the unfolded protein response. HAMLET activated all three main branches of the unfolded protein response and in addition a number of ER stress related genes were transcriptionally upregulated, including ATF4 and BIP. The activation of eIF2α in response to HAMLET may act as an “emergency break” to prevent further protein synthesis when the folding capacity of the ER is compromised. HAMLET treatment also triggered ATF6 cleavage and an increase in spliced XBP1, both acting to induce the transcription of a diverse set of chaperones and other ER-stress regulated genes, to augment the ER protein folding capacity. The inhibition by amiloride of the transcriptomic ER stress response and of eIF2α phosphorylation indicates that ion channel activation is an essential trigger of the ER stress response to HAMLET. In addition, the experiments reported hereinafter suggest that HAMLET interacts directly either with ER chaperones or the ER stress sensors as HAMLET has been shown to interact directly with proteasomes, which play a crucial role in ER stress and the unfolded protein response. HAMLET may also indirectly perturb the protein folding capability of carcinoma cells by decreasing ATP levels and by causing mitochondrial damage and permeabilization.
In healthy differentiated cells, HAMLET targeted innate immune signaling pathways involved in innate immunity and transiently suppressed p38 signaling. Although this immune response was low or absent in tumor cells, similar immune response pathways were strongly regulated in carcinoma cells under p38-specific inhibition, implying that these are not completely separate cellular response strategies. The mobilization intracellular Ca2+ by HAMLET in both carcinoma cells and healthy cells might indicate that this activation mechanism is shared, though of different magnitude. But the subsequent ion channel response was mainly observed in the carcinoma cells, which, however, suggests that this is the critical step to trigger cell death, a hypothesis also supported by the rescue effects of the ion channel blockers. This innate immune response would ideally serve to activate macrophages and other cells that scavenge and digest the remnants of apoptotic cells at sites of tissue damage and provide a suitable immune environment for cancer cell removal.
We speculate that the ability to selectively kill a broad range of tumor cells combined with the innate immune response of healthy differentiated cells gives rise to low toxicity in clinical studies as well as other beneficial effects. The p38 effector response in tumor cells accompanied by a beneficial innate immune response in surrounding tissue may serve as a two-tiered approach to killing cancer cells while maintaining tissue integrity.
The identification of the ion channel repertoire opens up a range of specific therapeutic options that will be expected to provide enhanced cancer therapies. Investigation of that repertoire has allowed the applicants to determine specific elements that can give rise to new therapeutic actives that form an aspect of the invention as described further below.